The prior art, in particular U.S. Pat. Nos. 4,344,840 and 4,457,829, describes processes for treating heavy hydrocarbon cuts comprising a first treatment step carried out in the presence of hydrogen in a reactor containing an ebullated bed of catalyst followed by a second step of fixed bed hydrotreatment. The descriptions illustrate the case of fixed bed treatment in the second step of a light gas fraction of the product from the first step.
One of the aims of the present invention is to produce, from certain particular hydrocarbon fractions which will be defined in the following description, by means of partial conversion of those fractions, lighter fractions which can easily be upgraded such as middle distillates (engine fuels: gasoline and diesel) and base oils.
Within the context of the present invention, the degree of conversion of the feed into lighter fractions is normally in the range 10% to 75%, or even 100% when the unconverted heavy fraction is recycled, and is usually in the range 25% to 60%, or even limited to 50%.
The feeds treated in the present invention are straight run vacuum distillates, vacuum distillates from a conversion process such as those from coking, from fixed bed hydroconversion such as those from HYVAHL.RTM. processes for treating heavy fractions developed by the Applicant, or from ebullating bed heavy fraction hydrotreatment processes such as those from H-OIL.RTM. processes, or from solvent deasphalted oils, for example propane, butane or pentane deasphalted oils originating from deasphalting a straight run vacuum residue or vacuum residues from HYVAHL.RTM. or H-OIL.RTM. processes. The feeds can also be formed by mixing those various fractions in any proportions, in particular deasphalted oil and vacuum distillate. They can also contain light cycle oil (LCO) of various origins, high cycle oil (HCO) of various origins and diesels from catalytic cracking generally having a distillation range of about 150.degree. C. to about 370.degree. C. They can also contain aromatic extracts and paraffins obtained from the manufacture of lubricating oils.
The aim of the present invention is to produce good quality products in particular with a low sulphur content under relatively low pressure conditions, to limit the necessary investment costs. This process can produce a gasoline type engine fuel containing less than 100 ppm by weight of sulphur thus satisfying the most severe specifications as regards sulphur content for this type of fuel, and this can be achieved using a feed which may contain more than 3% by weight of sulphur. Similarly, and this is particularly important, a diesel type engine fuel is obtained with a sulphur content much lower than 500 ppm and a residue with an initial boiling point which is, for example, about 370.degree. C. which can be sent as a feed or part of a feed to a conventional catalytic cracking step or to a residue catalytic cracking reactor such as a double regeneration reactor, preferably to a conventional catalytic cracking reactor.
We have now discovered, and this constitutes one of the aims of the present invention, that it is possible in the second step to treat either the whole of the product from the first ebullating bed conversion step, or the liquid fraction from that step, recovering the gas fraction converted in that first step, under favourable conditions leading to good stability of the ensemble of the system and improved selectivity for middle distillate.
In its broadest aspect, the present invention is defined as a process for converting a hydrocarbon fraction with a sulphur content of at least 0.3%, normally at least 1% and usually at least 2% by weight and with an initial boiling point of at least 300.degree. C., normally at least 340.degree. C. and usually at least 360.degree. C., and an end point of at least 400.degree. C., normally at least 450.degree. C. and which may reach 600.degree. C. or even 700.degree. C., characterized in that it comprises the following steps:
a) treating the hydrocarbon feed in a treatment section in the presence of hydrogen, said section comprising at least one three-phase reactor, containing at least one ebullating bed of hydroconversion catalyst the mineral support of which is at least partially amorphous, functioning in riser mode for liquid and for gas, said reactor comprising at least one means (5) located near the bottom of the reactor for extracting catalyst from said reactor and at least one means (4) located near the top of said reactor for adding fresh catalyst to said reactor; PA0 b) sending at least a portion, normally all, of the effluent from step a) to a section for treatment in the presence of hydrogen, said section comprising at least one reactor containing at least one fixed bed of hydrotreatment catalyst the mineral support of which is at least partially amorphous, under conditions for producing an effluent with a reduced sulphur content and a higher middle distillates content.
Normally, the treatment section of step a) comprises one to three reactors in series and the treatment section of step b) also comprises one to three reactors in series.
In a preferred implementation of the invention, at least a portion, normally all, of the effluent obtained from step b) is sent to a distillation zone [step c)] from which a gas fraction, a gasoline type engine fuel fraction, a diesel type engine fuel fraction and a liquid fraction which is heavier than the diesel type fraction are recovered.
In a variation, the heavier liquid fraction of the hydroconverted feed from step c) is sent to a catalytic cracking section ]step d)] in which it is treated under conditions for producing a gas fraction, a gasoline fraction, a diesel fraction and a slurry fraction.
In a further variation, at least a portion of the heavier fraction of the hydroconverted feed from step c) is sent either to ebullated bed hydroconversion step a), or to fixed bed hydrotreatment step b), or to each of these steps. It is also possible to recycle all of that fraction.
The gas fraction obtained in steps c) or d) normally principally comprises saturated and unsaturated hydrocarbons containing 1 to 4 carbon atoms per molecule (for example methane, ethane, propane, butane, ethylene, propylene, butylenes). At least a portion, preferably all, of the gasoline type fraction obtained in step c) is sent to the gasoline pool, for example. At least a portion, preferably all, of the diesel type fraction obtained in step c) is sent to the gasoline pool. At least a portion, preferably all, of the slurry fraction obtained from step d) is usually sent to the heavy fuel pool of the refinery, generally after separating out fine particles which it contains in suspension. In a further implementation of the invention, at least a portion, preferably all, of this slurry fraction is returned to the inlet to catalytic cracking step d). In a still further implementation of the invention, at least a portion of this slurry fraction is sent either to step a), or to step b), or to each of these steps, generally after separating out the fine particles it contains in suspension.
One particular implementation of the present invention comprises an intermediate step a1) between step a) and step b) in which the product from step a) is split into a heavy liquid fraction and a lighter fraction which is recovered. In this implementation of the present invention, the heavy liquid fraction obtained in step a1) is then sent to hydrotreatment step b). This implementation improves upgrading of the light fractions obtained from hydrotreatment step a) and limits the quantity of product to be treated in step b). The lighter fraction obtained in step a1) can be sent to a distillation zone from which a gas fraction, a gasoline type engine fuel fraction, a diesel type engine fuel fraction and a liquid fraction which is heavier than the diesel fraction are recovered at least part of which can, for example, be returned to step a) and/or at least part of which can be returned to converting hydrotreatment step b). The distillation zone in which this lighter fraction is split can be distinct from the distillation zone of step c), but usually this lighter fraction is sent to the distillation zone of step c).
In a particular implementation, which may be a preferred implementation when the catalyst used in step a) tends to form fines which can eventually alter the operation of the fixed bed reactor of step b), it is possible to provide a separation step b1) to eliminate at least a portion of the fines before introducing the product from either step a) or step al) into hydrotreatment step b). This separation can be carried out using any means which is known to the skilled person. As an example, separation can be carried out using at least one centrifuging system such as a hydrocyclone, or at least one filter. The scope of the present invention encompasses direct separation of the product from step a) then sending the product which is depleted in fines to step al), but this would involve treating a larger quantity of product than if separation were to be carried out on the liquid fraction from step al) when it exists. In a particular implementation of step b1), at least two separation means are used in parallel, one being used to carry out separation while the other is purged of retained fines.
The conditions of step a) for treating the feed in the presence of hydrogen are normally the conventional ebullating bed conditions for hydroconversion of a liquid hydrocarbon fraction. An absolute pressure of 2 to 35 MPa, normally 5 to 20 MPa and usually 6 to 10 MPa is used with a temperature of about 300.degree. C. to about 550.degree. C., normally about 350.degree. C. to about 500.degree. C. The hourly space velocity (HSV) and partial pressure of hydrogen are important factors which are selected depending on the characteristics of the product to be treated and the desired conversion. The HSV is usually in a range of about 0.1 h.sup.-1 to about 10 h.sup.-1, preferably about 0.5 h.sup.-1 to about 5 h.sup.-1. The quantity of hydrogen mixed with the feed is usually about 50 to about 5000 normal cubic metres (Nm.sup.3) per cubic metre (m.sup.3) of liquid feed, usually about 100 to about 1000 Nm.sup.3 /m.sup.3, preferably about 300 to about 500 Nm.sup.3 /m.sup.3. A conventional granular hydroconversion catalyst may be used which comprises at least one metal or metal compound with a hydro-dehydrogenating function on an amorphous support. This catalyst may be a catalyst comprising group VIII metals, for example nickel and/or cobalt, usually combined with at least one group VIB metal, for example molybdenum and/or tungsten. A catalyst may be used which, for example, comprises 0.5% to 10% by weight of nickel, preferably 1% to 5% by weight of nickel (expressed as nickel oxide NiO) and 1% to 30% by weight of molybdenum, preferably 5% to 20% by weight of molybdenum (expressed as molybdenum oxide MoO.sub.3) on an amorphous mineral support. The support can, for example, be selected from the group formed by alumina, silica, silica-aluminas, magnesia, clays and mixtures of at least two of these minerals. The support can also comprise other compounds, for example oxides selected from the group formed by boron oxide, zirconia, titanium oxide, and phosphoric anhydride. Usually, an alumina support is used, mainly an alumina support doped with phosphorous and possibly with boron. The concentration of phosphoric anhydride P.sub.2 O.sub.5 is normally less than about 20% by weight, usually less than about 10% by weight. This P.sub.2 O.sub.5 concentration is normally at least 0.001% by weight. The concentration of boron trioxide B.sub.2 O.sub.3 is normally about 0 to about 10% by weight. The alumina used is normally a .gamma. or a .eta. alumina. This catalyst is usually in the form of extrudates. The total concentration of oxides of group VIB and VIII metals is normally about 5% to about 40% by weight, generally about 7% to 30% by weight, and the weight ratio of the group VIB metal(s) to the group VIII metal(s), expressed as the metal oxide, is generally about 20 to about 1, usually about 10 to about 2. Used catalyst is partially replaced by fresh or new catalyst at regular intervals for example, in batches or quasi-continuously by extracting it from the bottom of the reactor. Fresh catalyst can be introduced every day, for example. The rate of replacing used catalyst with fresh catalyst can, for example, be about 0.05 kilograms to about 10 kilograms per cubic metre of feed. Extraction and replacement are carried out using apparatus which enable this hydroconversion step to be carried out continuously. The unit normally comprises a recirculation pump which maintains the catalyst in an ebullating bed by continuously recycling at least a portion of the liquid extracted from the head of the reactor and re-injecting it at the bottom of the reactor. It is also possible to send used catalyst extracted from the reactor to a regeneration zone in which the carbon and sulphur contained in the catalyst are eliminated, then to send the regenerated catalyst to hydroconversion step a).
Usually, this hydroconversion step a) is carried out under T-STAR.RTM. process conditions as described, for example, in the article "Heavy Oil Hydroprocessing", published by l'Aiche, Mar. 19-23, 1995, HOUSTON, Tex., paper number 42d. It can also be carried out under the conditions of the H-OIL.RTM. process as described, for example, in the article published by NPRA Annual Meeting, Mar. 16-18, 1997, J. J. Colyar and L. I. Wilson, entitled "THE H-OIL.RTM. PROCESS: A WORLDWIDE LEADER IN VACUUM RESIDUE HYDROPROCESSING".
The products obtained during step a) in the variation mentioned above [step a1)] are sent to a separation zone from which a heavy liquid fraction and a lighter fraction can be recovered. This heavy liquid fraction normally has an initial boiling point of about 350.degree. C. to about 400.degree. C., preferably about 360.degree. C. to about 380.degree. C., for example about 370.degree. C. The lighter fraction is normally sent to a separation zone in which it is split into light gasoline and diesel fractions at least part of which can be sent to gasoline pools, and into a heavier fraction.
In hydrotreatment step b), a conventional hydrotreatment catalyst is normally used, preferably at least one of those which have been described by the Applicant, in particular one of the catalysts described in European patents EP-B-0 113 297 and EP-B-0 113 284. An absolute pressure of about 2.5 to 35 MPa is normally used, usually about 5 to 20 MPa and more usually about 6 to 10 MPa. The temperature in step b) is normally about 300.degree. C. to about 500.degree. C., usually about 350.degree. C. to about 450.degree. C., and more usually about 350.degree. C. to about 420.degree. C. This temperature is normally adjusted depending on the desired level of hydrodesulphuration. The hourly space velocity (HSV) and partial pressure of hydrogen are important factors which are selected depending on the characteristics of the product to be treated and the desired conversion. The HSV is usually in a range of about 0.1 h.sup.-1 to about 5 h.sup.-1, preferably about 0.5 h.sup.-1 to about 2 h.sup.-1. The quantity of hydrogen mixed with the feed is normally about 100 to about 5000 normal cubic metres (Nm.sup.3) per cubic metre (m.sup.3) of liquid feed, usually about 200 to about 1000 Nm.sup.3 /m.sup.3, and preferably about 300 to about 500 Nm.sup.3 /m.sup.3. The operation is preferably carried out in the presence of hydrogen sulphide and the partial pressure of hydrogen sulphide is normally about 0.002 times to about 0.1 times, preferably about 0.005 times to about 0.05 times, the total pressure. In the hydrodesulphuration zone, the ideal catalyst must have a high hydrogenating power, so as to carry out thorough refining of the products, and to obtain a large reduction in sulphur. In the preferred implementation, the hydrotreatment zone operates at a relatively low temperature, which tends towards thorough hydrogenation and limits to coking. The scope of the present invention encompasses using a single catalyst or several different catalysts in the hydrotreatment zone, either simultaneously or successively. This step b) is normally carried out on an industrial scale in one or more reactors with a downflow of liquid.
In the hydrotreatment zone [step b)], at least one fixed bed of conventional hydrotreatment catalyst is used, the support of which is preferably at least partially amorphous. Preferably, a catalyst is used the support of which is, for example, selected from the group formed by alumina, silica, silica-aluminas, magnesia, silicamagnesias, clays and mixtures of at least two of these minerals. The support can also comprise other compounds, for example oxides selected from the group formed by boron oxide, zirconia, titanium oxide and phosphoric anhydride. Normally, an alumina support is used, usually an alumina support doped with phosphorous and possibly boron. The concentration of phosphoric anhydride P.sub.2 O.sub.5 is normally less than about 20% by weight, usually less than about 10% by weight. This P.sub.2 O.sub.5 concentration is normally at least 0.001% by weight. The concentration of boron trioxide B.sub.2 O.sub.3 is normally about 0 to about 10% by weight. The alumina used in normally a .gamma. or a .eta. alumina. This catalyst is usually in the form of extrudates or beads. A conventional granular hydrotreatment catalyst can be used which comprises at least one metal or metal compound having a hydro-dehydrogenating function, on an amorphous support. This catalyst can be a catalyst comprising group VIII metals, for example nickel and/or cobalt, usually combined with at least one group VIB metal, for example molybdenum and/or tungsten. As an example, a catalyst can be used which comprises 0.5% to 10% by weight of nickel, preferably 1% to 5% by weight of nickel (expressed as nickel oxide NiO), and 1% to 30% by weight of molybdenum, preferably 5% to 20% by weight of molybdenum (expressed as molybdenum oxide MoO.sub.3), on an amorphous mineral support. The total content of group VI and VIII metal oxides is usually about 5% to about 40% by weight, in general about 7% to 30% by weight, and the weight ratio, expressed as the metal oxide, between the group VIB metal (or metals) and the group VIII metal (or metals) is generally about 20 to about 1, usually about 10 to about 2.
In the distillation zone of step c), the conditions are generally selected so that the cut point for the heavy feed is about 350.degree. C. to about 400.degree. C., preferably about 360.degree. C. to about 380.degree. C., for example about 370.degree. C. In this distillation zone, a gasoline fraction with an end point which is usually about 150.degree. C. and a diesel fraction with an initial boiling point which is usually about 150.degree. C. and an end point of about 370.degree. C. are recovered.
Finally, in a variation mentioned above, in a catalytic cracking step d) at least a portion of the heavy fraction of the hydrotreated feed obtained in step c) can be sent to a conventional catalytic cracking section in which it is conventionally catalytically cracked under conditions which are well known to the skilled person to produce a fuel fraction (comprising a gasoline fraction and a diesel fraction) at least a portion of which is normally sent to gasoline pools, and a slurry fraction at least a portion, preferably all, of which is sent to a heavy fuel pool, for example, or at least a portion, preferably all, of which is recycled to catalytic cracking step d). Within the context of the present invention, the expression "conventional catalytic cracking" encompasses cracking processes comprising at least one partial combustion regeneration step and those comprising at least one total combustion regeneration step and/or those comprising both at least one partial combustion step and at least one total combustion step. In a particular implementation of the invention, a portion of the diesel fraction obtained during this step d) is recycled either to step a), or to step b) or to step d) mixed with the feed introduced into catalytic cracking step d). In the present description the term "a portion of the diesel fraction" means a fraction of less than 100%. The scope of the present invention encompasses recycling a portion of the diesel fraction to step a), a portion to step b) and a further portion to step d), the ensemble of these portions not necessarily adding up to the whole of the diesel fraction. It is also possible to recycle all of the diesel obtained by catalytic cracking either to step a), or to step b) or to step d), or a fraction to each of these steps, the sum of the fractions representing 100% of the diesel fraction obtained in step d). At least a portion of the gasoline fraction obtained in catalytic cracking step d) can also be recycled to step d).
A summary description of catalytic cracking (the first industrial use goes back to 1936 (HOUDRY process) or to 1942 for the use of a fluidised bed of catalyst) will be found in ULLMANS ENCYCLOPEDIA OF INDUSTRIAL CHEMISTRY VOLUME A 18, 1991, pages 61 to 64. A conventional catalyst is normally used, comprising a matrix, an optional additive, and at least one zeolite. The quantity of zeolite is variable but is normally about 3% to 60% by weight, usually about 6% to 50% by weight and more usually about 10% to 45% by weight. The zeolite is normally dispersed in the matrix. The quantity of additive is normally about 0 to 30% by weight, usually about 0 to 20% by weight. The quantity of matrix represents the complement to 100% by weight. The additive is generally selected from the group formed by oxides of metals from group IIA of the periodic table such as magnesium oxide or calcium oxide, rare earth oxides and titanates of group IIA metals. The matrix is usually a silica, an alumina, a silica-alumina, a silica-magnesia, a clay or a mixture of two or more of these products. The most frequently used zeolite is Y zeolite. Cracking is carried out in a substantially vertical reactor either in riser or in dropper mode. The choice of catalyst and the operating conditions are functions of the desired products depending on the feed treated as described, for example, in the article by M. MARCILLY, pages 990-991, published in the Institut Fran.cedilla.ais du Petrole review, November-December 1975, pages 969-1006. The temperature is normally about 450.degree. C. to about 600.degree. C. and the residence times in the reactor are less than 1 minute, usually about 0.1 to about 50 seconds.
The catalytic cracking step d) can also be a fluidised bed catalytic cracking step, for example using the process known as R2R developed by the Applicants. This step can be carried out in conventional fashion as known to the skilled person under conditions suitable for cracking with a view of producing lower molecular weight hydrocarbon products. Descriptions of the operation and catalysts for use in this context of fluidised bed cracking in this step d) are described, for example, in patents U.S. Pat. No. 4,695,370, EP-B-0 184 517, U.S. Pat. No. 4,959,334, EP-B-0 323 297, U.S. Pat. Nos. 4,965,232, 5,120,691, 5,344,554, 5,449,496, EP-A-0 485 259, U.S. Pat. Nos. 5,286,690, 5,324,696 and EP-A-0 699 224, the descriptions of which are hereby incorporated into the present description by reference.
The fluidised bed catalytic cracking reactor can operate in riser or dropper mode. Although it is not a preferred implementation of the present invention, it is also possible to carry out catalytic cracking in a moving bed reactor. Particularly preferred catalytic cracking catalysts are those containing at least one zeolite normally mixed with a suitable matrix such as alumina, silica, or silica-alumina.
In a particular implementation, when the treated feed is a vacuum distillate from vacuum distillation of an atmospheric distillation residue of a crude oil it is advantageous to recover the vacuum residue to send it to a solvent deasphalting step f) from which an asphalt fraction is recovered and a deasphalted oil is recovered at least part of which, for example, is sent to hydroconversion step a) mixed with the vacuum distillate. In this particular implementation, at least a portion of the slurry fraction obtained in catalytic cracking step d) can advantageously be recycled to the inlet to this deasphalting step f)
Solvent deasphalting step f) is carried out under conventional conditions which are well known to the skilled person. Reference should be made in this respect to the article by BILLON et al published in 1994 in volume 49, number 5 of the Institut Fran.cedilla.ais du Petrole review, pages 495 to 507, or to the description given in the description in French patent FR-B-2 480 773, or to the description in the Applicant's patent FR-B-2 681 871, or to the description of our patent U.S. Pat. No. 4,715,946, the descriptions of which are hereby incorporated by reference. Deasphalting is normally carried out at a temperature of 60.degree. C. to 250.degree. C. with at least one hydrocarbon solvent containing 3 to 7 carbon atoms, possibly with the addition of at least one additive. Suitable solvents and additives have been widely described in the documents cited above and in patent documents U.S. Pat. Nos. 1,948,296, 2,081,473, 2,587,643, 2,882,219, 3,278,415 and 3,331,394, for example. It is also possible to recover solvent using an opticritical process, i.e., using a solvent under supercritical conditions. This process can in particular substantially improve the overall economics of the process. Deasphalting can be carried out in a mixer-settler or in an extraction column. In the context of the present invention, a technique using at least one extraction column is preferred.
In a preferred implementation of the invention, the residual asphalt obtained from step f) is sent to an oxyvapogasification section in which is transformed into a gas containing hydrogen and carbon monoxide. This gas mixture can be used to synthesise methanol or to synthesise hydrocarbons using the Fischer-Tropsch reaction. In the context of the present invention, this mixture is preferably sent to a shift conversion section in which it is converted to hydrogen and carbon dioxide in the presence of steam. The hydrogen obtained can be used in steps a) and b) of the process of the invention. The residual asphalt can also be used as a solid fuel or, after fluxing, as a liquid fuel, or it can form part of a bitumen composition.
FIGS. 1, 2, 3 and 4 schematically represent the principal variations for implementing the process of the invention. In the Figures, similar items are designated by the same reference numbers and letters.
In FIG. 1, the hydrocarbon feed to be treated enters via line 10 into vacuum distillation zone (5) at the outlet from which a vacuum distillate is recovered via line 15 and a vacuum residue is recovered via line 11 which is sent to a solvent deasphalting zone (6) from which an asphalt is recovered via line 12 and a deasphalted oil is recovered via line 14 and optionally sent via line 13 to the vacuum distillate which leaves the distillation zone via line 15, which distillate is sent to a section (1) for treatment in the presence of hydrogen, said hydrogen being introduced via line 19. Catalyst is added via line 16 and extracted via line 17. The effluent treated in zone (1) is sent via line 18 to a zone (2) for hydrotreatment in the presence of hydrogen (introduced via line 19a); a portion of the treated effluent is optionally recovered via line 20 and a portion is sent via line 21 to a distillation zone from the outlet of which a gas fraction is recovered via line 22, a gasoline fraction is recovered via line 24 and a diesel fraction and a liquid fraction which is heavier than the diesel type fraction via line 25 which is optionally sent to hydrotreatment section (2) via line 26 and optionally to hydrotreatment section (1) via line 27, and is sent to a catalytic cracking section (4) via line 28, at the outlet from which a gas fraction is recovered via line 30, a gasoline fraction is recovered via line 31, a diesel fraction is recovered via line 29 and a slurry fraction is recovered via line 32, part of which is sent to the heavy fuel pool via line 33, a further portion of the slurry fraction optionally being sent via line 34 then via line 36 to the catalytic cracking section (4), a further portion of the slurry fraction optionally being sent via line 35, then via line 26 to the hydrotreatment section (2). A portion of the gasoline fraction is optionally sent via line 37 and line 36 to the catalytic cracking zone (4). A portion of the diesel fraction is optionally sent via line 38 then via line 36 to the catalytic cracking zone (4), a further portion of that fraction optionally being sent via line 39 to the hydrotreatment section (1), a further portion of that same fraction optionally being sent via line 40 to the hydrotreatment section (2).
In a particular embodiment of the invention described in FIG. 2, the effluent recovered via line 18 after treatment in section (1) is sent to a zone (1a) in which the effluent is split into a heavy fraction which is sent to hydrotreatment section (2) via line 18a and into a light fraction which is recovered via line 18c, a portion of that light fraction optionally being sent via line 18b to a distillation zone (3).
In a further particular embodiment of the invention described in FIG. 3, the effluent recovered via line 18 after treatment in section (1) is sent via lines 18d and 18e to zones (2a) and (2b) for separating fines from which an effluent is recovered via line 18f which is sent to the hydrotreatment section (2).
In the still further particular embodiment of the invention described in FIG. 4, the effluent recovered via line 18 after treatment in section (1) is sent to a zone (1a) in which said effluent is split into a heavy liquid fraction which is sent via lines 18d and 18e to zones (2a) and (2b) for separating fines from which an effluent is recovered which is sent to hydrotreatment section (2) via line 18f, and into a light fraction which is recovered via line 18c, a portion of that light fraction optionally being sent to distillation zone (3) via line 18b.